Complete Guide to Capacity Management Software

The room booking system says the 200-seat lecture theatre is fully booked from nine until four. Walk past at eleven and you will count perhaps a third of the seats filled. Two floors down, students are sitting on the library floor because every study space is taken. Both situations are happening in the same building, on the same day, and the data the estates team is looking at shows neither of them.

This is the gap capacity management software exists to close. Not the gap between what you planned and what you booked, but the gap between what you booked and who actually turned up. If you have already read the introductory articles, you know the vocabulary. What follows is the part that matters when you are signing a purchase order: what the software needs to do, where the measurement goes wrong, and how to make the output drive decisions rather than dashboards.

Three jobs, one platform

Good capacity management software does three distinct jobs, and buyers often pay for all three while only needing one at the time of purchase. It helps to name them separately.

• Live control. Real time occupancy monitoring of a zone, with an alert when headcount reaches a predefined limit. This is the fire-safety, queue-management and event-day use case.
• Utilisation analysis. Space utilization software that aggregates weeks of counts to show which rooms, floors and desk banks are used, when, and by how many. This is the portfolio and lease-decision use case.
• Building control. Passing live building occupancy into HVAC, lighting, BMS and security so the systems respond to people rather than to a timetable. This is the energy and operations use case.

The sensing hardware is the same for all three. The difference lies in the software layer and the integrations. A public library that mainly needs to know when a reading room is full has very different requirements from a corporate occupier deciding whether to hand back two floors at lease break. Be explicit internally about which job you are buying for first, because that decides which platform features deserve scrutiny.

Why the sensor choice decides everything downstream

Every number the software produces is only as good as the count feeding it. Badge data counts entries but rarely exits, so by mid-afternoon it overstates occupancy. Wi-Fi association counts devices, not people, and misses anyone with Wi-Fi off. PIR motion detection tells you a room is not empty, which is a different statement from how many people are in it.

For zone-level capacity, the reliable approach is bidirectional counting at every opening into the zone, using an optical or 3D occupancy sensor mounted overhead. The word "every" is doing real work in that sentence. One propped-open fire door in summer and the zone count drifts within an hour. A desk occupancy sensor under each workstation answers a different question, namely whether that specific desk is in use, and the two technologies are usually combined rather than chosen between.

On accuracy, ask vendors for a contractual figure, not a brochure figure. Vemco Group's installations carry a contractual minimum of 96%, and typically achieve 98 to 99% where lighting, layout and movement patterns allow. The honest version of that sentence includes the conditions. Any supplier quoting a flat 99% with no caveat has either not installed in a glass-walled atrium with low winter sun, or is not telling you about it.

One further feature matters more in workplaces and campuses than in retail: staff exclusion. AI-based sensors can recognise and exclude cleaning, security and facilities staff from the count, so a floor that reads as busy at seven in the morning is not just the cleaning team passing through. Without this, utilisation reports for early and late periods are quietly inflated.

The alert threshold is not the legal maximum

A common mistake when configuring an occupancy monitoring system is to set the alert at the licensed capacity of the space. By the time the alert fires, the people already in the queue outside are still going to enter, and the person receiving the alert has to physically get to the door. In practice the threshold should sit comfortably below the limit, with a second escalation level at the limit itself. The right gap depends on how fast the space fills and how far away the responder is, which is why it should be configurable per zone, not global.

Consider also who receives the alert and how. A screen at the entrance that turns from green to amber handles most situations without anyone being paged. SMS or app notifications to a duty manager should be reserved for the escalation level, otherwise alert fatigue sets in within a fortnight and the alerts get muted.

From counting to smart building automation

The largest recurring saving from capacity management rarely comes from the alerts. It comes from stopping the building running at full output for people who are not there. Most HVAC schedules are set once, to the longest plausible occupied day, and never revisited. A floor that empties at three on Fridays is heated and ventilated until seven.

Vemco's VemFusion connects live occupancy data to HVAC, BMS and security systems so that ventilation rates, setpoints and lighting zones follow actual headcount. Universities and public buildings use this specifically to cut the waste from always-on systems in spaces with irregular use, such as lecture theatres, sports halls and council chambers. The data for utilisation analysis comes from the same sensors through VemSpace, so the energy case and the space case are built on one installation rather than two.

Two practical points for anyone taking this to the building services team. First, the BMS integration should run on a published protocol or API, and the sensor platform should expose the count at zone level, not just whole-building. Second, start with ventilation and lighting on a single floor and compare against the previous year's consumption before extending. Facilities teams who promise building-wide savings in the first quarter tend to find the estate team is already sceptical.

What implementers know that the brochure does not say

Here is the observation that comes up on nearly every workplace project. In the first weeks after sensors go live, the workplace occupancy analytics almost always show lower use than the booking data, and someone senior will question whether the sensors are wrong. Resist recalibrating. Walk the floors at the times in question and count by hand. The hand count will match the sensors, and the real finding is that booked space and used space have been diverging for years with nobody measuring it. Plan for that conversation before the dashboard goes live, because it lands better when it is expected.

The second thing implementers learn is that occupancy management software is only trusted when the counts are visible to the people being counted. Universities that publish live library occupancy on screens and in the student app see fewer complaints and better spread across study spaces, because students route themselves to where the seats are. Transparency does more for adoption than any internal report.

Building the business case

Finance will ask for one number. Give them three lines instead, each with its own evidence trail.

• Space. Measured peak utilisation across the portfolio against current cost per square metre. This is where the lease and consolidation decisions live, and it needs at least one full term or quarter of data.
• Energy. Consumption on a pilot floor before and after occupancy-driven control, normalised for weather.
• Compliance and operations. Hours of manual counting removed, incidents avoided, and the audit trail for licensed capacity that the software produces automatically.

The space line is usually the largest, the energy line is the easiest to prove, and the compliance line is the one that gets the project approved by risk and estates together. Sequence the pilot so that the energy evidence arrives first.

Frequently asked questions

What is an occupancy sensor? An occupancy sensor is a device that detects the presence or number of people in a defined space and reports that data to a software platform. Simple versions only detect motion; counting sensors distinguish individuals and direction of travel, which is what capacity management requires.

How do occupancy sensors work? Overhead counting sensors use optical or 3D depth imaging to recognise people passing beneath them and record whether each one entered or left. The software sums entries and exits across every opening into a zone to maintain a live headcount, with AI models filtering out staff, trolleys and other non-visitor movement so the number reflects actual occupancy.

If you are weighing a capacity management project for an office, campus or public building and want to see how zone counting, alert thresholds and BMS integration would be scoped for your specific spaces, talk to the Vemco Group team about a pilot floor and what the first quarter of data would tell you.